Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
re Welding
=,.$
%
%
SINCE
BI =
CaF,
Chipman1 and Belton et a13 found an effective equilibrium temperature for the MnO and the SiO2reactions
respectively in selected fluxes, to be 2000 OC based
upon chemical analysis of the weld metal.
Perhaps more important than either of these justifications is the finding which will be presented subsequently that equilibria based upon 2000 OC agree with
the experimental data while equilibria based upon other
temperatures do not.
Assumption 2. Kubli and Sharav2 showed that the
oxygen content of submerged arc weld metal decreases
with increasing flux basicity. Tuliani et al,I8quantified
the oxygen content of the weld metal as a function of
the basicity index. The International Institute of Welding flux basicity index (BI) used by Tuliani et a1 is given
by:
0L
I
2
Flux B a s i c i t y Index
I
3
'
- 28360
log K = -+ 10.61
T
K = - fisiO^
[21
asto,
where the standard state for SiO, in the slag is pure SiO,
and the standard state for oxygen and silicon in the
weld metal is based upon a 1 pet solution.
2. Calculation Procedure
Belton el a13have derived the following thermodynamic data for the Si02reaction in the temperature
range 17 13 to 2000 'C.
Table
System
Temperature('C)
Which
Reference
Flux Basicity I n d e x
Fig. 3-The relationship between the equilibrium percent SiO, and
the flux basicity index for a given weld metal Si content in the
SO2-MnO-FeOsystem at 2000 'C.
0.5
1.0
F l u x B a s i c i t y Index
1.5
Fig. 2-The relationship between the activity of SiO, and the flux
basicity index for a given weld metal Si content at 2000 "Cassuming
regular solution behavior in the flux.
METALLURGICAL TRANSACTIONS B
F l u x Basicity I n d e x
Fig. &The relationship between equilibrium percent MnO and the
flux basicity index for a given weld metal Mn content in the
MnO-CaO-Al,03system at 2000 OC. The dotted lines are presented
for reference to discussion in the text.
log K =
12760
- 5.68
T
0.8
-1.2
-30
-20
-10
0
A MnO,%
10
20
T= 2000C
30
161
and
AMnO
[71
MnO, - MnO,,eq.
Where Mn, is the measured final weld metal manganese content, Mn, is the average initial electrodel
baseplate manganese content, MnO, is the initial flux
MnO content, and MnO,,^ is the percent MnO in the
slag which would be in equilibrium with the final weld
metal manganese as predicted by Fig. 4.
If in the example given Mn, = 1.0 pet, then AMn
= 1.0 - 0.5 = 0.5 pet and from Fig. 4 we find AMnO
= 4.0 - 1.8 = 2.2 pet. AMnO is a measure of the
deviation from equilibrium and may be considered as a
stimulus while AMn is the response to this stimulus. If
the initial flux contains an excess of MnO with respect
to the initial flux basicity and electrode/baseplate
equilibrium, MnO from the slag will decompose thereby
increasing the weld metal Mn content. The larger the
initial MnO deviation from the final predicted equilibriuim value, the greater is the stimulus for MnO
decomposition. As noted above, AMn is a measure of
the response to this stimulus. Reversing the order of the
initial-final subscripts between Eq. [6] and [7], ensures
that a positive stimulus will produce a positive response
and conversely a negative stimulus will provide a
negative response. In a physical sense, Eqs. [6] and [7]
are written such that a loss of manganese from the slag
as predicted by the theory (positive) will result in a gain
of manganese in the weld metal (also positive). If a
positive stimulus should produce a negative response, or
vice versa, the equilibrium predicition of the theory
would be in error.
Such tests of the theory may be made if one has the
following information:
a) The initial flux composition. Equation [l] is then
I
0
0.5
1. 5
-"
mc
S 0.I
n
-.-VI"
c
.-
*0'
0
;-0.1
I
0.1
0.2
0.4
0.6
0.8
1.0
F l u x Basicity Index
1.2
0.2
0.3
0.4
0.5
Initial Silicon C o n t e n t , percent
0.6
Table Ii. Comparison of Theoretically Predicted Mn and Si Equilibrium with the Values Measured Experimentally
Flux Designation
Type
F- 1
CaO-MnO-SiO,
Flux Basicity
Index
Predicted
Mn, Pet
Measured
Mn, Pet
Predicted
Si Pct
Measured
Pct
Reference
1.17
1.83
1.85
0.33
0.28
This work
This work
F-4
CaO-SiO,
MIT-2
CaO-SiO,
0.46
1.1
1.27
1.45
0.52
0.2
0.15
0.15
0.52
<0.32
1.3
<0.5
0.80
0.95
This work
0.32
0.25
This work
0.25
This work
>1.18
This work
F-c
CaO-Si0,-Al,O,
F-d
CaO-SiOTAl,O,
F-e
CaO-MnO-SiO,
F-f
CaO-MnO-SiO,
n.d.--not determined.
All fluxes except F-1, F-2, MIT-1 and MIT-2 are commercial compositions.
Test
Number
Initial
Manganese
Mn,i, Pct
Final Surface
Manganese
Mn,.s, Pet
Final Bulk
Manganese
Mn,b, Pet
Final Bulk
Oxygen
0,b, pet
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1.85
1.40
0.09
1.74
1.00
0.85
0.07
0.94
0.98
0.64
1.95
0.88
1.15
1.26
1.08
1.10
1.01
1.11
1.25
0.96
1.85
1.85
0.82
1.75
1.50
1.62
1.50
1.90
1.50
0.36
0.36
0.33
0.53
0.39
0.29
0.44
0.39
1.85
1.50
0.48
1.75
1.15
1.20
0.40
I .27
I .40
1.00
1.90
1.10
0.64
0.56
0.55
0.65
0.69
0.57
0.60
0.60
0.050
0.047
0.068
0.072
0.052
0.062
0.035
0.030
0.049
0.060
0.046
0.056
0.086
0.077
0.068
0.084
0.071
0.053
0.063
0.090
Surface
Equilibrium
Temperature
T.,'C
Bulk
Equilibrium
Temperature
T,,'C
1979
1968
1896
203 1
1950
1992
1885
1979
1932
1812
203 1
1905
1941
1694
1835
1928
1906
1968
1910
2164
2115
2088
2158
2138
2047
2088
2158
1968
1962
2050
203 1
1994
21 19
203 1
1929
203 1
2074
544-VOLUME
METALLURGICAL TRANSACTIONS B
will be noted that none of the data falls within either the
II or IV quadrants. This confirms that the stimulus
predicted by the equilibrium theory agrees with the
response as determined experimentally. Considering the
imprecision with which one knows the actual weld
metal chemistry as well as the imprecision of the
thermodynamic data, the agreement of the theory and
experiment is quite remarkable.
It is worthy of note, that a similar analysis using
thermodynamic data based upon 1800 O C would not
yield good agreement, while an analysis based upon an
equilibrium temperature of 2200 'C, compresses the
data so as to obscure any trends. In addition, weld
metal equilibrium reaction temperatures based upon the
SiO-,reaction (Eq. [2]) agree with the 2000 OC assumption as may be seen in Fig. 7. Several of the welds of
Fig. 7 were preoduced with 3 pet silicon iron baseplate,
using very high basicity (low silica activity) fluxes. In
such a case one would expect transport of a large
amount of silicon from the weld pool to the slag. The
finding that the weld/slag equilibrium remains at
2000 OC suggests that silicon transport is very rapid
during submerged arc welding. The data confirm the
finding of Belton et aI3 that the SiO, reaction achieves
an effective equilibrium reaction temperature of
2000 OC over a wide range of weld metal chemistry and
flux composition.
- 0.2
0.2
0.4
A S i , percent
Fig. 12-The predicted reltionship between ASi and ASiO, for
commercial flux L91. The prediciton indicates little possible change in
weld silicon content for initial values between 0.3 and 0.5 pet silicon.
0.5
1.5
I 1
I
2
2.5
0.1
0.~2-~-0.3 0.4
0.5
0.6
0.7
- 0.2
0.2
Si
04
0.6
0.8
, percent
A Mn, percent
Fig. 14Th predicated relationship between AMn and AMnO for
commercial flux L60.
METALLURGICAL TRANSACTIONS B